Anthropic’s Claude Opus 4.6 AI Model Identifies 22 Security Vulnerabilities in Mozilla Firefox

Anthropic’s Claude Opus 4.6 AI model identified 22 security vulnerabilities in Mozilla Firefox in January 2026, the majority of which were classified as high severity. Mozilla addressed all discovered flaws in Firefox version 148, which was released in the same month. The vulnerabilities were found through automated analysis conducted by Anthropic’s AI system, though no specific CVE IDs or technical exploit details were disclosed. The report confirms the bugs were fixed but does not describe their individual impacts or attack vectors. The discovery highlights the role of AI-driven security research in identifying software vulnerabilities.